ID results - new model

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Simon Harrison
Posts: 19
Joined: Mon Oct 29, 2007 5:39 pm

ID results - new model

Post by Simon Harrison » Mon May 25, 2009 11:46 pm

Hi,

I have a model of the equine forelimb that I have set up for inverse dynamics. I have populated the inertial matrices with values that produce good results for sagittal plane analyses that I have carried out in Matlab. However when I run with either external loads specified or not specified, I get the following results for each state:

1.#QNAN000000000000000

Do you have any troubleshooting tips? I am happy with the inertial properties. No inertial properties are huge or too small.

All joints are hinges except for the most proximal joint which has 6 DoF. Could this be the cause of the error?

User avatar
Ayman Habib
Posts: 2235
Joined: Fri Apr 01, 2005 12:24 pm

RE: ID results - new model

Post by Ayman Habib » Wed May 27, 2009 2:25 pm

Simon,

I don't think the types of joints would be a problem, neither the dynamics of the multibody system since these are all used by inverse dynamics.
Do you have muscles in your model? If so, how are you initializing them, or setting their properties?

-Ayman

User avatar
Simon Harrison
Posts: 19
Joined: Mon Oct 29, 2007 5:39 pm

RE: ID results - new model

Post by Simon Harrison » Wed May 27, 2009 4:34 pm

Ayman

I think my post may have been confusing - I have carried out simplified ID calcs in Matlab successfully, but when I try to implement in OpenSim I get "1.#QNAN000000000000000" in each state entry.

I tried removing all muscles and the results don't change.

Regards,
Simon

User avatar
Simon Harrison
Posts: 19
Joined: Mon Oct 29, 2007 5:39 pm

RE: ID results - new model

Post by Simon Harrison » Wed May 27, 2009 11:23 pm

I think I have solved the problem. By de-constructing the model down to one joint and back again I found two causes:

1. One inertial value was reasonably small and when modified I was able to extract results. Note that this value was not excessively small.
2. I had a locked joint in the model. I have removed this joint and have connected the adjacent segments through a modified joint.

Now I can get ID to output numerical results.

POST REPLY